Background Neonatal acute kidney injury (AKI) has been associated with unfavorable outcomes, including increased mortality. We aimed to describe the clinical course and outcomes during the first 7 days after diagnosis in newborns with AKI in three neonatal intensive care units in Popayán-Colombia. Methods Multi-center prospective cohort study conducted between June 2019 and December 2020 in three NICUs after ethical approval. We included newborns between 2 and 28 days of life, first diagnosed with AKI using the KDIGO classification modified for newborns which consider increased serum creatinine values over baseline values as well as urine output over time in hours or both. Patients with chromosomal abnormalities, major kidney malformations, and complex congenital heart disease were excluded. Patients were followed for up to 7 days after diagnosis and the maximum KDIGO stage, recovery of kidney function, need for renal replacement therapy and cumulative incidence of death were evaluated. Results Over the 18 months of the study, 4132 newborns were admitted to the NICUs, and 93 patients (2.25, 95% CI 1.82–2.75%) developed neonatal AKI. 59.1% of the newborns were premature and there were no differences in severity according to gestational age. During follow-up, the maximum KDIGO was 64.5% for AKI-stage 1, 11.8% for AKI-stage 2, and 23.7% for AKI-stage 3. Kidney function recovery was higher in AKI-stage 1 patients vs. AKI-severe (AKI-stage 2 and 3) (95% vs. 48.5%). Five patients (5.4%) received renal replacement therapy and 15 died (16.1%), four in AKI-stage 1 vs. 11 in AKI-severe (6.7% vs 33.3%). Conclusions Newborns admitted to the NICUs can develop AKI regardless of gestational age, and it is more frequent between the second and ninth days of life. More patients whit AKI-stage 1 recover and die less than those in a severe stage.

Primary membranous nephropathy (MN) is an autoimmune disease caused by autoantibody binding to podocytes in human glomeruli. The disease leads to end-stage renal disease in about 30% of patients, even after treatment with immunosuppressive therapies. The discovery of serum autoantibodies against the Phospholipase A 2 Receptor 1 (PLA 2 R1) in 70% of patients with MN has resulted in substantial improvements of diagnosis, disease classification, treatment and prognosis. However, due do the lack of a rodent animal model, it has not been formally proven that human PLA 2 R1-antibodies induce disease. With the aim to define the role of human PLA 2 R1-antibodies in MN pathogenesis, we studied the impact of human PLA 2 R1-antibodies in minipigs. Methods: For passive transfer experiments, minipigs received plasma or purified total IgG from patients with PLA 2 R1-associated MN or from healthy controls. Further, active immunization was performed using human recombinant PLA 2 R1 protein. The presence of PLA 2 R1antibodies and development of proteinuria were monitored in the serum and urine of minipigs. Kidney samples were analyzed using immunohistochemistry, immunofluorescence and electron microscopy. Results: PLA 2 R1 is expressed on the podocytes on the minipig in a pattern identical to that in human kidneys. Human PLA 2 R1-antibodies bind to the minipig PLA 2 R1 both in vitro and in vivo. The passive transfer experiments using plasma or purified IgG from patients with PLA 2 R1-associated MN to minipigs resulted in development of all functional and morphologic characteristics of human MN in minipigs. Human PLA 2 R1-antibodies were detectable in the blood of the minipigs and could be eluted from the glomeruli of the minipigs. Importantly, the eluted human PLA 2 R1-antibodies were of the IgG4 subclass. Transfer of healthy human plasma of total IgG did not lead to any histomorphological changes resembling MN. The active immunization of minipigs with the human PLA 2 R1 protein led to generation of anti-PLA 2 R1 antibodies, which also resulted in the induction of all characteristics of human MN in the kidneys of the minipigs. Conclusions: By applying a translational approach from humans to minipigs we demonstrated that human PLA 2 R1-antibodies are pathogenetic and induce MN. In addition, we created a humanized model of autoimmune PLA 2 R1-induced MN by active immunization of minipigs with PLA 2 R1.